N1SRP — broadcast
20260920 033000 UTC
421 transcript segments
Google Speech-to-Text API Automatic Transcription (Chirp)
Data courtesy of The GDELT Project (https://www.gdeltproject.org/), from the Internet Archive TV News Archive.
Machine transcription. Treat it as a searchable index of what was broadcast, not a verbatim quotation record.

[00:00:00] It's somewhere even those machines can't reach. Researchers are testing a
[00:00:04] very different kind of first responder, cockroaches equipped with
[00:00:08] mini sensors. These biobots can crawl through tiny gaps and
[00:00:12] hazardous areas looking for survivors. And
[00:00:15] of rescue doesn't only happen on land. Underwater, strong currents and poor
[00:00:19] visibility can create new challenges. That's where
[00:00:23] underwater drones come
[00:00:24] in. This compact
[00:00:28] underwater drone is designed to... make the work of rescue teams
[00:00:32] of easier, according to its developer, it can be
[00:00:36] deployed in about 10 minutes and operate a depth of up to
[00:00:40] 300 m, and its hor
[00:00:43] battery technology supposedly gives it a wired operating
[00:00:47] range of about 2
[00:00:48] of km for up to 4 hours.
[00:00:52] we
[00:00:52] of focus really on the most challenging
[00:00:54] of conditions,
[00:00:55] i so here in Switzerland for example we have rivers which are quite challenging to
[00:00:59] to operate in - humans underwater drones are used to
[00:01:02] locate water ammunitions, about 1.6 million tons were
[00:01:06] dumped in the north and Baltic Seas during the World Wars, posing
[00:01:10] long-term environmental and safety risks. Using
[00:01:14] HD cameras and acoustic sensors or Sona, the robot can survey large
[00:01:18] areas either autonomously or under remote
[00:01:21] a control. Once it identifies an object, operators can guide it for
[00:01:24] a closer inspection and
[00:01:26] of recover loads of up to 40
[00:01:31] of of kilos. Using AI and sensor of fusion technology, the drone
[00:01:33] in combines information from cameras and Sona, allowing it to navigate even
[00:01:37] in merky
[00:01:37] water.
[00:01:42] You have both basically visual information and acoustic
[00:01:45] of information uh displays this a 3D
[00:01:47] of map um in real time for the operator, or of course if you have an
[00:01:50] autonomous interface, it's much simpler you just say, well we need to search
[00:01:54] in that area or inspect that kind of structure and it does it for you
[00:01:58] basically. The system has already been used in search and
[00:02:02] rescue missions. It could also support offshore wind, oil and gas
[00:02:05] inspections, reducing the need for dives in hazardous
[00:02:09] conditions. Drones like this take
[00:02:13] years of engineering and
[00:02:14] of lot of money, but not every AI
[00:02:17] of rescue tool starts
[00:02:18] u in a big company or research lab. This wildfire detection system
[00:02:21] began with an idea from teenager who wanted to protect his
[00:02:25] community. Today it's helping detect wildfires
[00:02:29] around the world. "I started in 2020 because I personally had
[00:02:33] family friends who lost their homes to wildfires and I really
[00:02:37] wanted to do something about it. Franco Rodriguez Viao
[00:02:40] of was 16 when he co-founded satellites on fire, an AI powered
[00:02:44] early warning system for wildfires. Today company figures show the
[00:02:48] of platform is used in more than 20 countries and helped authorities respond to
[00:02:52] more than 600 wildfires in 2025. The company
[00:02:56] also says its alert system can detect fires faster than NASA's.
[00:03:02] We combine satellites, cameras, drones, sensors and AI
[00:03:06] to detect fires. We have an AI algorithm that is
[00:03:10] enabling us to detect much faster by combining different kinds
[00:03:14] of information
[00:03:15] in with reports from customers in
[00:03:18] of territory. The platform can send smartphone alerts to people who are in
[00:03:22] danger and notify first responders.
[00:03:25] Franco's goal is to save lives and to protect forests,
[00:03:29] wildlife. and fragile
[00:03:30] ecosystems.
[00:03:36] We want to be an integral solution for this problem globally, so this means
[00:03:39] being in prevention, detection, monitoring and suppression
[00:03:42] of the
[00:03:42] fires.
[00:03:46] of But many regions simply don't have this technology. In many
[00:03:50] places hit hardest by climate related disasters, rescue teams
[00:03:54] work with whatever they have. Pakistan is among the countries most
[00:03:57] vulnerable to climate change. every monsoon, floods kill hundreds
[00:04:01] and in the worst years force millions from their homes. the warning
[00:04:05] systems often don't reach everyone, that's where citizen lad
[00:04:09] technology can help. nothing but water
[00:04:13] where roads and fields won stood, pakistan's monsoon floods are deadly
[00:04:17] threat. in 2022 alone flooding killed more than 1,700 people
[00:04:21] and affected 33 million others, nearly one in seven
[00:04:24] Pakistanis. in some flood prone areas, early warning systems still have
[00:04:28] gaps, study's point to prof
[00:04:57] of the surface
[00:05:01] radar sensors work similarly with radio waves and according to the
[00:05:04] company remain reliable even in bad
[00:05:07] weather
[00:05:10] is solar radiation aapki.
[00:05:42] isn't easy, mobile coverage can be
[00:05:44] patchy, but where they work, the sensus
[00:05:48] give authorities more time to want people and
[00:05:51] evacuate,
[00:05:55] in karachi, millions of people live in informal settlements that are hard
[00:05:59] hit by flooding, many these areas are poorly mapped,
[00:06:03] making evacuation even more
[00:06:04] difficult. local initiative is now
[00:06:08] digitizing old maps and combining them with satellite imagery to
[00:06:30] we take all of that data together and we collectively turn it into a
[00:06:34] live map which lets you choose your mode of transport
[00:06:38] and based on that it tells you
[00:06:41] what is for you to take. For
[00:06:45] now, these tools remain local and limited, but as flads
[00:06:49] become more severe, the challenge is getting them to the people who need
[00:06:52] them
[00:06:53] most.
[00:06:57] One the best rescue tools is also one the simplest phone
[00:07:01] alerts. They can save more lives than the most expensive rescue
[00:07:04] equipment. When powerful earthquake struck Venezuela in June
[00:07:08] 2026, the country had no public early warning system.
[00:07:12] but millions of android phones still sounded alarm. Using
[00:07:16] data from smartphones, Google's alert system detected the quake and
[00:07:20] warned people before the strongest shaking arrived. Chile, by
[00:07:24] contrast, has a state run warning system, one the most established in
[00:07:28] Latin America. It sends alerts to millions of phones within
[00:07:31] seconds. At Chile's National Disaster
[00:07:35] Prevention and Response Service every second counts. The team
[00:07:39] issues early warnings and coordinates the response earth.
[00:08:12] technology, instead of sending millions of individual text messages,
[00:08:16] alerts are transmitted simultaneously from every cell tower in the danger
[00:08:20] zone to all compatible phones nearby, even if the networks are
[00:08:23] overloaded, the warning still gets
[00:08:25] through.
[00:08:29] telecommunications companies have expanded network coverage across the entire
[00:08:33] territory, this allies us to
[00:08:36] reach even very remote areas with alerts and effectively cover the
[00:08:40] whole country.
[00:09:12] minutos: an emergency, those extra minutes can make all the
[00:09:16] difference. so here's the takeaway: the
[00:09:20] most effective disaster technology isn't always the most
[00:09:23] advanced, it's the technology that reaches people when they need it
[00:09:27] most. fast, reliable phone alerts are the first line of
[00:09:31] defense. AI, drones and robots are helping rescue
[00:09:35] teams go even
[00:09:48] Sve što treba da znate na jednom mestu u pola
[00:09:52] sata N1
[00:09:54] dnevnik.
[00:12:36] has said he agrees
[00:12:37] with you. Samtman says he agrees with you. What's the action
[00:12:40] plan for
[00:12:41] you guys to come together? Is
[00:12:42] there a group chat that's going to come into fruish in here? So so first of all, are you
[00:12:46] talking? I'm enormously appreciative
[00:12:50] that other industry leaders including our
[00:12:53] competitors have. have offered their
[00:12:56] agreement with this plan, i you, i think,
[00:12:59] i'm i'm very glad that uh they did this, and
[00:13:03] i think it's an important first step. um, are you
[00:13:07] guys talking? think i'll get to that. i think the next
[00:13:10] step is for multiple industry players, some them have already
[00:13:14] said they're going to do this to to also um
[00:13:18] have to also agree to have third party
[00:13:21] evaluators. then i think the next step. after
[00:13:25] that is in some form the
[00:13:29] companies need to work together to set
[00:13:32] standards about safety standards for release,
[00:13:36] perhaps something about the pace of release, but in order to do
[00:13:40] that, there there needs to be involvement the government,
[00:13:44] because what we're talking about is the rate of release of
[00:13:47] products because of whether they're safe, and in order to have that
[00:13:51] conversation, we need to have that conversation with the...
[00:13:55] government in the room, the conversation right now is that we
[00:13:59] are in the precipice of something really scary, is this
[00:14:03] five alarm fire? i you know, i don't,
[00:14:06] depends what you mean by five alarm fire. what what i would say is how would you describe
[00:14:10] it? what what i would describe is so something i've been
[00:14:14] saying since uh you know since we started anthropic, almost since before we
[00:14:18] started anthropic is that this technology is on an exponential
[00:14:22] right, and that's a mathematical function that you know... doubles every six months
[00:14:26] or doubles every three months or something like that, and so when something
[00:14:29] doubles every small amount of time,
[00:14:33] it compounds very quickly, right? it's like you have one, then two, then
[00:14:36] four, then eight, then 16, then 32,
[00:14:39] and as I was saying that, what I was trying to say is this
[00:14:43] technology, it's going to jump much more quickly
[00:14:47] than we imagine, and but even as I was
[00:14:51] saying that to myself, I don't think I fully just appreciate...
[00:14:55] what it would actually be like when the progress was as fast as it was,
[00:14:59] so we're we're an exponential curve looks a little like this. where are we on the curve?
[00:15:03] we're kind of we're kind of on the bend where it's starting to get steeper. so is
[00:15:06] that five alarm fire to you? what, it's what it what it
[00:15:10] means, it it doesn't mean we need to panic today,
[00:15:14] it doesn't mean we need to shut it all down. what i
[00:15:18] i would say, it's it's a warning sign, it's a warning
[00:15:22] sign that we need to slow down we. need to make this
[00:15:26] technology carefully and we need to make sure
[00:15:29] that the our safeguards our ability to
[00:15:33] understand it, our ability to control it keeps
[00:15:36] up. with the pace at which the technology
[00:15:40] is
[00:15:40] happening.
[00:15:47] yeah, i think that they're really serious. it's a sign that these people
[00:15:51] are noticing that the recent instance um,
[00:15:55] are much worse
[00:15:56] than lot
[00:15:56] of people were expecting, and that we need to do something pretty serious to handle them and make sure that
[00:16:00] something terrible doesn't happen. yes, generally speaking,
[00:16:04] one way of making an AI not do something bad is to put. rails around it so that
[00:16:08] even if it tries it can't break out the box or
[00:16:10] something like that uh, but the problem is that this kind of approach doesn't scale very
[00:16:14] well as the AI is get more and more capable. alignment is about making the
[00:16:18] AI uh inherently want to pursue the goals that we wanted to
[00:16:22] pursue or have the values that we wanted to have and that's the kind of thing that we need as they
[00:16:25] get more advanced. yeah, it's there's there's different ways it can go wrong, but
[00:16:29] at least one the ways it can go wrong is sort the like literal
[00:16:33] genie, you tell it to do something but you slightly misspecify.
[00:16:37] want what you wanted to do and it gets way out of hand. i think the hugging pase incident is an
[00:16:41] interesting example of this where the AIs were sort of trained to try and get a
[00:16:45] good score and they found out
[00:16:47] that the best way to get the good score involved breaking out onto the
[00:16:51] internet and hacking a public company. i think it's a hard question.
[00:16:54] um so definitely first
[00:16:56] of all amusingly open air was founded in similar way where sort of sam and elon musk were
[00:17:00] like we're not sure that we trust google to be making this stuff. so actually all the top
[00:17:04] competitors were born out of not trusting each other. i think... that there is
[00:17:07] definitely some component them spinning this news a way that is more
[00:17:11] favorable to them, but very fundamentally if they could snap their fingers
[00:17:15] and have it be that their ais do not break out and randomly hack things, they would
[00:17:19] definitely want to do that, so I don't think it's all spin, I don't think it's all
[00:17:23] someilaberpon, I think they are genuinely scared, but you do have to
[00:17:27] take the exact phrasing with a pinch of salt. The honest answer, and this is kind of
[00:17:30] unfortunate is that AI is becoming more and more of a
[00:17:34] critical piece of national security infrastructure. and so I
[00:17:38] would guess that the US government had at least some part to play in
[00:17:42] the decision, the more it becomes the case that these things have you,
[00:17:46] for example, previously anthropic and open AI had their own internal programs to figure out
[00:17:50] who would get private access to these models for defense, now the US government has kind of
[00:17:54] taken over that, I think UKAC is absolutely wonderful, I think that some part of
[00:17:58] what's going on is the the UK government doesn't have as much leverage and the third
[00:18:02] party auditors they're looking for are very much us auditors, very difficult
[00:18:06] question um mean a couple of things, what have we seen
[00:18:10] lately? so i think one thing that people underrate is that the China is
[00:18:14] in some ways more sympathetic to slowing down, because they're
[00:18:17] aware the fact that if AI gets very powerful,
[00:18:20] it could be bad for the PRC, could disrupt their ability to sort
[00:18:23] of um function effectively as a sort of dominant
[00:18:26] government. um, so actually compared to the US,
[00:18:30] they have some appetite for it, they are also currently behind, so there's a lot of
[00:18:34] specifics in how you do anything like slowing down or
[00:18:37] Yeah,
[00:18:44] so they're not exactly open source in
[00:18:46] that what they release is the contents to run them yourself, but not to actually build them
[00:18:50] yourself or see how they've done, see how they build, like traditional open source
[00:18:53] software,
[00:18:54] you can seal the code, if it's safe, but that
[00:18:57] decision
[00:18:58] to to public things, it's very common in tech in general for people who
[00:19:02] are a little behind the frontier to want to make the things more available, so sort of google doing this
[00:19:06] with android was a... story and windows doing it compared to Apple by
[00:19:10] releasing sort of Windows OS. I'm certainly hoping
[00:19:14] that it ends up being a dam squib, but I think that the chances that it aren't are
[00:19:17] substantial enough that we need to take this very, very seriously. I mean, if you
[00:19:21] look at the kinds of people who are worried, it's a lot the most
[00:19:25] heavyweight, well respected AI research
[00:19:28] luminaries, and the predictions that the people who are most worried were
[00:19:32] making 10 years ago are beginning to come true, so definitely watch this
[00:19:36] space. Uspeh ne čeka, zato je tu Forbes
[00:19:40] Srbija mobilna aplikacija, tvoj dnevni izvor biznis ideja,
[00:19:43] inovativnih rešenja, kao i inspirativnih priča lidera,
[00:19:47] pristupi najvažnijim vestima, ekskluzivnim analizama,
[00:19:50] intervjuima, bilo kad, bilo gde, biznis, uspeh,
[00:19:54] Forbes.
[00:20:31] Ether was um uh the mad madatter the
[00:20:35] family to be honest she was the one that got. to mischief, susanna
[00:20:38] sister was the sensible one, um, but yeah, she
[00:20:42] was a lively, lively, fun, loving little
[00:20:45] girl, um, that you know would always make
[00:20:49] us all laugh, and fall about and would be up to mischief
[00:20:53] whenever the opportunity erose, not bad mischief, but
[00:20:56] enough to,
[00:20:57] you know, you had to be to keep you on your toe, yeah,
[00:21:01] yeah, yeah, and as any girl her age is
[00:21:04] encouraged to be, should be, she was active, she liked being outdoors,
[00:21:08] yeah, and she did that, lot of your
[00:21:12] photographs and memories of her, yeah, she loved the beach, she loved the beach, she loved
[00:21:16] being outdoors, yeah, she collected hundreds and hundreds of shells,
[00:21:20] 'and we had to keep them all and and it was only she could
[00:21:24] only collect them if she promised to return some and only take
[00:21:27] few home um but of course you never got to choose was
[00:21:31] this family trip to Devon that that changed the course of your
[00:21:35] family's life and ultimately
[00:21:37] led to the end of heathers can you talk to me about what happened that day because it was an ordinary
[00:21:41] family trip to the beach yeah yeah she she was enjoying herself
[00:21:45] having lots of fun and we were going to go for a walk on the evening
[00:21:48] and um previously in the day we'd done the
[00:21:52] same walk along the seafront along the path um and
[00:21:56] that's where she trod in that uh the the water from the combined sewage
[00:22:00] overflow we carried on the day she played on the beach did
[00:22:04] normal things but then in the evening when we went for our evening
[00:22:07] walk uh she started she suddenly became
[00:22:10] unwell um and and became
[00:22:14] quite distressed uh so we got her back to
[00:22:18] the chalet and did what ever parent would do and go you got a b
[00:22:21] sickness and diarrhea, you can't have anything eat or drink, you'll be better
[00:22:25] soon, don't worry about it, I'll stay with you, dad and Suzanne will go
[00:22:29] out, um, but she she didn't, she
[00:22:33] wasn't getting better, um, so we took it to the
[00:22:37] chemist, chemist prescribed to hiroly, you know, sickness and die,
[00:22:40] same, every, every parent's journey with a child that's got,
[00:22:44] sickness and diarrhea, um, but then as things
[00:22:48] progress she started to bleed, and that's when we went back to the doctors
[00:22:52] and that's when we were rush to casualty um and everything you saw
[00:22:56] in dirty business is what happened um which is the the channel
[00:23:00] 4 series yeah based on your
[00:23:04] family's experience and exploring the state the UK waters
[00:23:07] and the the industry involved in it southwest water
[00:23:11] carried out investigations at the time um despite formal
[00:23:15] investigations no cause for the outbreak was identified um
[00:23:19] the inquest patient confirming that no definitive source of infection was established and the
[00:23:23] circumstances southwest water say of waste water infrastructure
[00:23:26] regulation the late 1990s is very different from today and
[00:23:30] since then there has been significant investment. however,
[00:23:34] can I just say something? yeah,
[00:23:36] you can. At the time it took them 30 days to
[00:23:40] take samples from the water in the area, and we're always told that
[00:23:44] if there's
[00:23:44] a sewage spillage into the sea, 48 hours waiting
[00:23:48] to see, but they... did not, they assumed at the time,
[00:23:52] they completely assumed it was from food, and when they drew a
[00:23:56] blank, that's when they started looking at the environment.
[00:24:00] when you think about it, heather died back in 1999
[00:24:04] and water company say you're a very different place to to where
[00:24:07] we were in the 1990s, but are we that? well, that's
[00:24:11] the thing, I mean, you went to Westminster with
[00:24:15] surface against Sewage last week with some very
[00:24:19] start. statistics from the summer
[00:24:22] period, um, also with personal
[00:24:26] stories from families, this summer, there was a young girl called Keeva, eight years
[00:24:29] old, she became seriously ill after swimming at Aven Beach in
[00:24:33] Doorset again, designated bathing water site
[00:24:36] awarded excellent, she swallowed seawater while swimming at her
[00:24:40] local beach and was hospitalized afterwards,
[00:24:43] what is it like for you to see this happening up
[00:24:47] and down the country?
[00:24:50] repeatedly, what is it like for you as family who have lost a
[00:24:53] daughter? it it's a mixture, there's
[00:24:57] a anger, real anger, but also there's
[00:25:01] a deep sadness,
[00:25:04] because when i had my meeting with the secretary of state and the
[00:25:08] secretary for water, the minister for water, sorry, i said
[00:25:12] to them really clearly, i am not going to
[00:25:15] stop until something changes, because i can't
[00:25:20] not allow this to happen to another child and another
[00:25:23] family, i can't, because it's
[00:25:26] horrific, and you can't, you can't ignore the
[00:25:30] fact that this is a public health crisis, the the
[00:25:34] suwage system, it's primary function is to protect
[00:25:38] public health, and it's not doing
[00:25:40] that, and so I just cannot understand why
[00:25:44] there are all these people in power, all these people that have the
[00:25:48] authority, they must... children, they must have
[00:25:51] families, they must feel,
[00:25:54] you know, they must understand, and yet they continue
[00:25:58] to do the same things over and over again, so...
[00:26:02] is discharging sewage into bathing water, why would you do
[00:26:06] that? why would you run the risk on your
[00:26:09] watch, the possibility of another death? why would you
[00:26:13] do that? final question, julie, um, water safety became a
[00:26:17] mandatory part the correctrium in the UK in schools
[00:26:21] to do with water safety, we had an unprecedented number of deaths this year from
[00:26:25] people getting into difficulty open water swimming, it didn't include
[00:26:29] water quality, we know that that does, get taught
[00:26:33] as part of geography lessons, but in your case you actually brought
[00:26:37] our guide with the charity service against sewage to make people
[00:26:41] aware, because that's the thing, people go to the seaside and they
[00:26:45] don't always have an app to check what the quality of water is like,
[00:26:48] but you want to educate young people, because while they're
[00:26:52] perhaps not being looked after by the state or water companies, they need to
[00:26:55] look after themselves, what do you want to see in schools? well, the
[00:26:59] people that are taking them on the trips, doing the risk assessments, do a full risk
[00:27:03] assessment, looking at water quality, including that in
[00:27:06] the risk assessment, and also making children aware. I mean, how many,
[00:27:10] I mean I know, I cuz I used to work in education, if kids got dirty, just
[00:27:14] rinse your hands in the water when you're on an outdoor ed, just give your hands and it's
[00:27:18] like, now that's not the case, that is not a safe place to actually
[00:27:22] make sure you're clean, you know, use some wet wipes, use antiback,
[00:27:25] use something, don't touch your lip, you know, all that kind of basic
[00:27:29] stuff, don't touch your lips, wash your hands before e
[00:27:33] hygiene, think the water as potentially
[00:27:36] contaminated, don't assume it's not, buse
[00:27:40] we can't assume that the water companies, the environmental
[00:27:44] agency, the government are looking after us at the
[00:27:47] moment, they aren't, so we have to protect ourselves
[00:27:50] and that's my big message at the moment is don't
[00:27:54] assume, take care, look after your children, keep them
[00:27:58] safe, the water isn't safe.
[00:28:09] Od 12. do 18. septembra, svake večeri
[00:28:13] u 9:00 na
[00:28:15] Sinemaniji.
[00:28:19] Stivendorf triler drami
[00:28:22] dug. Jamie Bell
[00:28:25] u romantičnoj drami, filmske zvezde ne umiru u
[00:28:29] Liverpulu.
[00:28:34] Daniel Redcliff u akcionoj komediji
[00:28:37] Pištoljdžija
[00:28:44] Natali Portman u akcionoj Western drami Opasna
[00:28:47] Jain
[00:28:48] David
[00:28:52] Eliot ratnoj drami Kadžaki
[00:28:56] istinita
[00:28:56] priča
[00:29:01] Colin Farel. U drami ubistvo svetog
[00:29:04] Jelena,
[00:29:11] Sharon Stone u romantičnoj drami, sve što
[00:29:14] želim, više filmova,
[00:29:18] viševa.
[00:29:20] Sinemanija.
[00:29:25] U novom broju radara Miodrag Jovanović, izbori idu u pravcu referendumske
[00:29:29] atmosfere, izborne liste, karte se otvaraju, šta je Niklo na Zgarištu?
[00:29:32] palate na pepelu, Srbija i BiH diplomatskom Klinču Mladić,
[00:29:36] rizična izborna platforma, gde nestaje novac iz budžeta, asfalt važniji od
[00:29:40] penzionera.
[00:29:46] U novom danu tražimo odgovore na pitanja koja su važna za sve
[00:29:49] nas. Samo dobro informisani, možete donositi prave
[00:29:53] odluke.
[00:30:27] Dobrodošli, poštovanje, gledate subotnji pregled sport klub
[00:30:31] vesti, put je. od odbojke do fudbala, na početku gledamo vesti
[00:30:35] dana. Još dva dana do Slovenije,
[00:30:39] Srbija se sprema za osminu finala Evropskog
[00:30:42] prvenstva.
[00:30:47] Bryton deklasirao Arsenal u derbiju petog kola premijer
[00:30:50] lige.
[00:30:54] Harry Kein, rekorder Bundeslige, 100 golova za
[00:30:57] 98 mečeva.